JP2022071982A - Oscillator - Google Patents

Abstract

To propose a configuration that suppresses oscillation at a frequency higher than the oscillation frequency fosc in a microstrip resonator such as a patch antenna.SOLUTION: An oscillator that oscillates terahertz waves includes a negative resistance element including a first semiconductor layer, a second semiconductor layer, and an active layer provided between the first semiconductor layer and the second semiconductor layer. A first conductor, a second conductor, and a dielectric provided between the first conductor and the second conductor constitute a resonator. The negative resistance element is provided between the first conductor and the second conductor. A layer having a resistivity higher than that of the first semiconductor layer or the second semiconductor layer, or an amorphous layer is provided between the negative resistance element and the dielectric.SELECTED DRAWING: Figure 1

Description

The present invention relates to an oscillator for a high frequency electromagnetic wave (referred to as a terahertz wave in the present specification) having an arbitrary frequency band from the millimeter wave band to the terahertz wave band (30 GHz or more and 30 THz or less).

In the terahertz wave band, absorption peaks derived from the structure and state exist for many organic molecules such as biomaterials, pharmaceuticals, and electronic materials. In addition, terahertz waves have high permeability to materials such as paper, ceramics, resins, and cloth. In recent years, research and development of imaging technology and sensing technology utilizing such characteristics of terahertz waves have been carried out. For example, it is expected to be applied to a fluoroscopic inspection device instead of an X-ray device, an in-line non-destructive inspection device in a manufacturing process, and the like.

As a current injection type oscillator that generates electromagnetic waves in this frequency band, an oscillator in which a resonator is integrated in a negative resistance element is well known.

Patent Document 1 discloses a terahertz wave oscillator in which a negative resistance element, which is a resonant tunneling diode (RTD), and a microstrip resonator are integrated on the same substrate.

It is known that an oscillator using a negative resistance element causes parasitic oscillation caused by a bias circuit including a power supply and wiring for adjusting the bias voltage of the negative resistance element. Parasitic oscillation refers to parasitic oscillation in a frequency band on the low frequency side different from the desired frequency, and reduces the oscillation output at the desired frequency.

Patent Document 1 proposes a configuration that suppresses parasitic oscillation. FIG. 9 shows the structure of Patent Document 1.

9 (a) is a perspective view showing the appearance of the oscillator 1100, and FIG. 9 (b) is a cross-sectional view thereof. The oscillator 1100 is a microstrip resonator having a patch antenna 1102 and configured to sandwich a negative resistance element 1101 and a dielectric 1107 between two conductors 1108 and 1109. The resonance frequency is fosc.

The bias circuit for adjusting the bias voltage of the negative resistance element 1101 is composed of a power supply 1105 and a wiring 1106. Since the wiring 1106 always has a parasitic inductance component, it is displayed as an inductance. The strip conductor 2031 supplies a bias to the negative resistance element 101 from the bias circuit. The resistor 1104 and the capacitance 2032 connected in parallel with the resistor 1104 are low impedance circuits. These suppress the parasitic oscillation of fsp1 (fsp1 <fosc, typically the frequency band from DC to 10 GHz) which is a relatively low frequency caused by a bias circuit such as a power supply 1105 or a wiring 1106.

Further, in FIG. 9, a resistance element 1110 arranged in parallel with the negative resistance element 1101 is provided in the patch antenna 1102. The inductance of the strip conductor 2031 and the capacitance of the patch antenna 1102 form an LC resonance with a frequency of fsp2 (fsp2 <fosc). The frequency fsp2 is determined by the length of the strip conductor 1031 and the area of the patch antenna 1102, and is typically in the range of 10 to 500 GHz. The resistance element 1110 is arranged in a substantial node of the electric field having an oscillation frequency of fosc. As a result, the impedance near the desired oscillation frequency fosc is high, and the impedance near the frequency fsp2 (fsp2 <fosc) of the parasitic oscillation is low, so that the parasitic oscillation due to the wiring structure can be suppressed.

Japanese Unexamined Patent Publication No. 2014-199965

By the way, it is generally known that in a system using an alternating current, harmonics having a frequency that is an integral multiple of the fundamental wave caused by a power source or an electronic device are generated with respect to the basic frequency.

Similarly, an oscillator using a negative resistance element often generates harmonics that are integral multiples of the desired oscillation frequency fosc, and the generation of harmonics reduces the oscillation output of the desired frequency fosc. There's a problem.

In the above-mentioned Patent Document 1, the oscillation fhr (fhr> fosc) of harmonics higher than the desired frequency fosc is not considered, and there is a problem that unnecessary harmonics cannot be suppressed.

An object of the present invention is to provide an oscillator capable of stably oscillating at a desired terahertz band oscillation frequency by suppressing oscillation of an unnecessary frequency higher than a desired oscillation frequency in a microstrip resonator such as a patch antenna. It is to be.

The oscillator according to the present invention is an oscillator that oscillates a terahertz wave, and is provided between the first semiconductor layer, the second semiconductor layer, the first semiconductor layer, and the second semiconductor layer. A resonance having a negative resistance element having an active layer, a first conductor, a second conductor, and a dielectric provided between the first conductor and the second conductor. The negative resistance element is provided between the first conductor and the second conductor, and a layer is provided between the negative resistance element and the dielectric. The layer is a layer having a resistance higher than that of the first semiconductor layer or the second semiconductor layer.

Further, the conductor according to the present invention is an oscillator that oscillates a terahertz wave, and is between the first semiconductor layer, the second semiconductor layer, the first semiconductor layer, and the second semiconductor layer. A negative resistance element having an active layer provided, a first conductor, a second conductor, and a dielectric provided between the first conductor and the second conductor. The negative resistance element is provided between the first conductor and the second conductor, and a layer is provided between the negative resistance element and the dielectric. Is provided, and the layer is an amorphous layer.

According to the present invention, in a microstrip resonator such as a patch antenna, an oscillator capable of stably oscillating at a desired terahertz band oscillation frequency by suppressing oscillation of an unnecessary frequency higher than a desired oscillation frequency is provided. can.

The figure explaining the structure of the 1st Embodiment of the oscillator which concerns on this invention. The figure explaining the structure of the 2nd Embodiment of the oscillator which concerns on this invention. The figure explaining the structure of the 3rd Embodiment of the oscillator which concerns on this invention. The figure explaining the structure of the 4th Embodiment of the oscillator which concerns on this invention. The figure explaining the structure of the 5th Embodiment of the oscillator which concerns on this invention. The figure explaining the structure of the 6th Embodiment of the oscillator which concerns on this invention. The figure explaining the structure of the oscillator 200 of Example 1. The figure explaining the structure of the oscillator 300 of Example 2. The figure explaining the structure of patent document 1.

(First Embodiment)
The oscillator 100 according to this embodiment will be described with reference to FIG. 1A and 1B are schematic views showing an oscillator 100 according to the first embodiment of the present invention, FIG. 1A is a perspective view showing an appearance, and FIG. 1B is a cross-sectional view taken along the line AA'.

The oscillator 100 of the present invention is a microstrip resonator configured such that a negative resistance element 101 and a dielectric 107 are sandwiched between a conductor 108 (first conductor) and a conductor 109 (second conductor). The negative resistance element 101 is provided between the conductor 108 (first conductor) and the conductor 109 (second conductor).

The structure of the negative resistance element 101 is a mesa structure that forms a trapezoid, a rectangle, a circle, or the like in a plan view. As shown in FIG. 1A, here, a mesa structure having a circular cross-sectional shape is shown. The negative resistance element 101 is composed of low resistance semiconductor layers 102a and 102b due to doping or the like, and an active layer 103 having a gain. That is, in the negative resistance element 101, the low resistance semiconductor layer 102a (first semiconductor layer), the active layer 103, and the low resistance semiconductor layer 102b (second semiconductor layer) are laminated in the first direction. ing.

A layer 110 having a higher resistivity than the low resistance semiconductor layers (doping layers) 102a and 102b is provided on the side wall of the negative resistance element 101. In the present embodiment, the layer 110 is divided into 110a, which is the side wall of the doping layer 102a, 110b, which is the side wall of the doping layer 102b, and 110c, which is the side wall of the active layer 103.

The layers 110a, 110b, 110c have a higher resistivity than the doping layers 102a and 102b. Specifically, the resistivity of the layers 110a, 110b, 110c is 10 times or more and 10000 times or less of the resistivity of the doping layer. For example, it is 100 times or more and 1000 times or less.

The layers 110a, 110b, 110c are amorphous layers 110a, 110b, 110c in which a layer in an amorphous state is formed by forming a negative resistance element 101 into a mesa shape and then performing a treatment process.

Here, the amorphous means a state having no clear crystallinity, but very fine crystals may be present in a part thereof. Specifically, it is a case where there is no diffraction peak showing crystallinity in the electron diffraction measurement by X-ray diffraction spectrum (XRD) or TEM observation. It is also the case that the peak is slight even if it exists. Further, if a halo peak is present, it is assumed to be amorphous.

The negative resistance element 101 including the active layer 103 is an element in which a region in which the current decreases as the voltage increases, that is, a region having a negative resistance appears in the current-voltage characteristics. The negative resistance element 101 is typically a high-frequency element such as a resonance tunnel diode (RTD), an esaki diode, a Gunn diode, or a transistor having one terminal terminated. For example, a tannet diode, an impat diode, a heterojunction bipolar transistor (HBT), a compound semiconductor electric FET, a high electron mobility transistor (HEMT), or the like may be used. Further, the negative resistance of the Josephson element using a superconductor may be used.

In the following, a case where RTD, which is a typical negative resistance element operating in the terahertz band, is used for the negative resistance element 101 will be described as an example.

The patch antenna 104 is a terahertz band resonator, and is a microstrip type resonator using a microstrip line having a finite length or the like. The patch antenna 104 is configured such that the negative resistance element 101 and the dielectric 107 are sandwiched between the two conductors 108 and 109. The conductor 108 is the upper conductor of the patch antenna 104, and the conductor 109 is the ground conductor of the patch antenna 104.

Here, the dielectric is a substance in which dielectric property is superior to conductivity, and is a material that behaves as an insulator or a high resistance body that does not conduct electricity with respect to a direct current voltage. Typically, a material having a resistivity of 1 kmΩm or more is suitable. Specific examples include plastic, ceramic, silicon oxide, silicon nitride, silicon oxynitride, and the like.

The patch antenna 104 is set so that the width of the conductor 108 in the AA'direction is a λ / 2 resonator. The patch antenna 104 is an active antenna in which a negative resistance element 101 is integrated. Therefore, the oscillation frequency fosc defined by the patch antenna 104 of the oscillator 100 is determined as the resonance frequency of the all-parallel resonant circuit in which the reactance of the patch antenna 104 and the negative resistance element 101 are combined. Specifically, Jpn. J. Apple. Phys. , Vol. 47, No. According to the disclosure of 6 (2008), the frequency satisfying the following conditions, which is a combination of the admittance of the RTD (Y _RTD ) and the admittance of the antenna (Y _ANT ), is the oscillation frequency fosc. Equation (1) is an amplitude condition, and equation (2) is a phase condition.
Amplitude condition Re [Y _RTD ] + Re [Y _ANT ] <= 0 (1)
Phase condition Im [Y _RTD ] + Im [Y _ANT ] = 0 (2)

Here, Re [Y _RTD ] is the admittance of the negative resistance element and has a negative value.

The bias circuit for adjusting the bias voltage of the negative resistance element 101 is composed of the power supply 105 and the wiring 106. Since the wiring always has a parasitic inductance component, it is shown as an inductance in FIG. The power supply 105 supplies the current necessary for driving the negative resistance element 101 and adjusts the bias voltage. The bias voltage is typically selected from the negative resistance region of the negative resistance element 101. The strip conductor 1031 has a role of supplying a bias from the power supply 105 and the wiring 106 to the negative resistance element 101.

The resistor 112 and the capacitance 1032 connected in parallel with the resistor 112 are low impedance circuits. This suppresses the parasitic oscillation of a relatively low frequency caused by the bias circuit such as a power source 105 and a wiring 106.

By the way, when a high-frequency current flows through a conductor, it is known that a larger current flows toward the surface of the conductor due to the skin effect, and the higher the frequency, the more the current concentrates on the surface.

Here, when the depth: d that becomes 1 / e of the current flowing on the surface of the conductor is called the skin depth, and the electrical resistivity: ρ, the magnetic permeability: μ, and the angular frequency: ω, d is given by the equation (3). ).
d = √ (2ρ / ωμ) (3)

The skin depth d becomes smaller as the frequency ω is higher, and the current passes only in the vicinity of the surface as the frequency ω is higher.

Since the decrease in the skin depth d is the decrease in the area through which the current passes, the higher the frequency, the higher the effective resistance value. As an approximate value of the skin depth in the terahertz band, for example, the skin depth when a high-frequency current flows through a low resistance semiconductor layer having a resistivity of 3 × 10 ^-7 Ωm is about 260 nm at 1 THz, whereas it is about 260 nm at 2 THz. It is about 160 nm.

Even in driving the mesa-shaped negative resistance element 101 that oscillates a terahertz wave, the higher the frequency due to the skin effect, the more current flows near the surface of the mesa, and the effective resistance value becomes higher.

As described above, in the oscillator 100 according to the present embodiment, the amorphous layers 110a and 110b are formed on the side walls of the doping layers 102a and 102b. With this configuration, in the oscillation circuit of the negative resistance element 101, a loss occurs due to the combined resistance of the doping layers 102a and 102b and the amorphous layers 110a and 110b. The higher the frequency due to the skin effect, the more the current passes near the surface of the mesa, so that the effective resistance value becomes higher and the loss becomes larger due to the influence of the amorphous layers 110a and 110b formed on the surface side. .. As a result, the second harmonic and third harmonic, which are higher frequencies than the oscillation frequency fosc of the oscillator 100, are more affected by the amorphous layer 110 and the loss becomes larger, so that unnecessary harmonics are oscillated. It becomes possible to suppress. The amorphous layers 110a and 110b can provide an effective loss inside the patch antenna 104 at a very close position on the side wall of the negative resistance element 101.

The resistivity of the amorphous layers 110a and 110b has a required oscillation frequency fosc value and negative resistance so that the loss of the oscillation frequency fosc of the oscillator 100 is small and the loss is effective for the second and third harmonics. It can be appropriately set according to the structure of the resistance element 101 and the like. The components constituting the amorphous layers 110a and 110b are mainly composed of elements constituting the doping layers 102a and 102b, elements in the process required for amorphization, and the like. For example, InGaAs, InAlAs, n-InGaAs, Si, O, N, C, F and the like can be mentioned.

The thicknesses of the amorphous layers 110a and 110b are appropriately set depending on the required oscillation frequency fosc, the structure of the negative resistance element 101, and the like, but a range of about 1 nm or more and 500 nm or less is appropriate. A more appropriate numerical value is in the range of 5 nm or more and 200 nm or less. Here, the thickness (film thickness) is the thickness (film thickness) in the direction (second direction) orthogonal to the direction (first direction) in which the layers constituting the negative resistance element 101 are laminated. Is.

Further, as described above, the amorphous layer 110c is formed on the side wall of the active layer 103. The components constituting the amorphous layer 110c on the side wall of the active layer 103 are mainly composed of components constituting the active layer 103, components of the steps required for amorphization, and the like. For example, undoped AlAs, InGaAs, InAlAs, and O, N, C, F, etc., which are not carrier-doped, can be mentioned. The thickness of the amorphous layer 110c is appropriately set depending on the required oscillation frequency fosc, the structure of the negative resistance element 101, and the like, but is generally in the range of 1 nm or more and 500 nm or less. A more appropriate numerical value is in the range of X5 nm or more and 200 nm or less.

The resistivity of the amorphous layer 110c on the side wall of the active layer 103 is higher than that of the doping layers 102a and 102b, and the impedance of the active layer 103, which is an undoped semiconductor, is lowered by amorphization. Therefore, the leakage current of the capacitance of the active layer 103 increases, and the dielectric loss tangent (tan δ) becomes large. Here, it is known that when an AC electric field is applied to a dielectric, a power loss (dielectric loss) occurs, and the dielectric loss is proportional to the frequency and the dielectric loss tangent (tan δ). Therefore, the second-order harmonics and the third-order harmonics, which have a higher frequency than the oscillation frequency fosc of the oscillator 100, have a larger dielectric loss and have an effect of suppressing the oscillation of unnecessary harmonics.

(Second embodiment)
2A and 2B are schematic views showing an oscillator 100 according to a second embodiment of the present invention, FIG. 2A is a perspective view showing an appearance, and FIG. 2B is a cross-sectional view taken along the line AA'.

In FIG. 2, the negative resistance element 101 has a low resistance semiconductor layer (doping layer) 102a and 102b and an active layer 103. The description of the same structure as that of the first embodiment will be omitted.

A layer 110 is formed on the side wall of the negative resistance element 101. The layer 110 in the second embodiment is formed by forming a negative resistance element 101 into a mesa shape and then forming a resistance film having a desired resistivity. As a component of the resistance film and a film forming method, a CVD method or a CVD method is used in which Si, O, N, F, P, S, etc. are reacted with a low resistance material such as Al, Ti, C so as to obtain a desired resistivity. It can be formed by using a sputter method, a vapor deposition method, or the like. For example, examples of the layer 110 include nitrides such as aluminum oxide, titanium oxide, silicon carbide, silicon nitride, titanium nitride, and aluminum nitride, and compounds such as sulfur-based compounds and fluorine-based compounds.

This configuration causes a loss due to the combined resistance of the doping layers 102a and 102b and the layer 110. As the frequency increases due to the skin effect, the current flows near the surface of the mesa, so that it is greatly affected by the layer 110 formed on the surface side, resulting in higher resistance and greater loss. As a result, the second harmonic and the third harmonic, which have a higher frequency than the oscillation frequency fosc of the oscillator 100, have a large loss due to the amorphous layer 110, so that it is possible to suppress the oscillation of unnecessary harmonics. Become.

(Third embodiment)
FIG. 3 is a schematic view showing the oscillator 100 according to the third embodiment of the present invention, and is a schematic cross-sectional view (a diagram corresponding to FIG. 1B) showing the periphery of the negative resistance element 101. The description of the same structure as that of the first embodiment will be omitted.

In FIG. 3, layers (amorphous layers) 110a and 110b are formed on the side walls of the low resistance semiconductor layers (doping layers) 102a and 102b of the negative resistance element 101. The amorphous layers 110a and 110b have higher resistivity than the doping layers 102a and 102b. Even in this configuration, as described above, the loss of the second harmonic and the third harmonic, which are high frequencies, is large with respect to the oscillation frequency _fosc of the oscillator 100, so that the oscillation of unnecessary harmonics can be suppressed.

Here, a current path between the active layer 103 is formed in the form of the doping layer 102a, the amorphous layer 110a on the side wall, and the doping layer 102b and the amorphous layer 110b on the side wall combined. Therefore, at least a part of the doping layer 102a and the amorphous layer 110a on the side wall, and the doping layer 102b and the amorphous layer 110b on the side wall may be electrically connected to the active layer 103. Therefore, the positions of the ends of the side walls of the active layer 103 and the ends of the amorphous layers 110a and 110b can be arbitrary.

(Fourth Embodiment)
FIG. 4A is a schematic view showing the oscillator 100 according to the fourth embodiment of the present invention, and is a schematic cross-sectional view (a diagram corresponding to FIG. 1B) showing the periphery of the negative resistance element 101. .. The description of the same structure as that of the first embodiment will be omitted.

In FIG. 4A, the amorphous layers 110a and 110b are formed on the side walls of the doping layers 102a and 102b of the negative resistance element 101. The side wall of the active layer 103 is closer to the center side of the mesa structure than the side wall of the amorphous layers 110a and 110b. FIG. 4B is a plan view, and is configured such that the outer edge of the active layer 103 is arranged closer to the center of the mesa structure than the outer edges of the amorphous layers 110a and 110b.

As in this configuration, the positions of the amorphous layers 110a and 110b and the positions of the active layer 103 can be arbitrarily set. Contrary to FIG. 4, the outer edges of the amorphous layers 110a and 110b may be configured to be located closer to the center of the mesa structure than the outer edges of the active layer 103.

(Fifth Embodiment)
FIG. 5 is a schematic view showing the oscillator 100 according to the fifth embodiment of the present invention, and is a schematic cross-sectional view (a diagram corresponding to FIG. 1B) showing the periphery of the negative resistance element 101. The description of the same structure as that of the first embodiment will be omitted.

In FIG. 5, the amorphous layer 110c is formed on the side wall of the active layer 103 of the negative resistance element 101. The resistivity of the amorphous layer 110c is higher than that of the doping layers 102a and 102b, and the impedance is low with respect to the active layer 103.

Therefore, as described above, the leakage current of the capacitance of the active layer 103 increases, and the dielectric loss tangent (tan δ) becomes large. As a result, the dielectric loss of the second harmonic and the third harmonic, which are high frequencies, increases with respect to the oscillation frequency fosc of the oscillator 100, so that there is an effect of suppressing the oscillation of unnecessary harmonics.

Since the active layer 103 and the amorphous layer 110c are connected to each other, the position of the outer edge of the amorphous layer 110c and the position of the outer edge of the doping layers 102a and 102b electrically connected to the active layer 103 can be formed into any shape. ..

(Sixth Embodiment)
FIG. 6A is a schematic view showing the oscillator 100 according to the sixth embodiment of the present invention, and is a schematic cross-sectional view (a diagram corresponding to FIG. 1B) showing the periphery of the negative resistance element 101. .. The description of the same structure as that of the first embodiment will be omitted.

In FIG. 6A, the amorphous layer 110c is formed on the side wall of the active layer 103 of the negative resistance element 101. In the plan view shown in FIG. 6B, the outer edges of the doping layers 102a and 102b are configured to be arranged closer to the center of the mesa structure than the outer edges of the amorphous layer 110c.

As described above, the positions of the outer edge of the amorphous layer 110c and the outer edges of the doping layers 102a and 102b can be arbitrarily set. Therefore, in a plan view, the outer edge of the amorphous layer 110c may be configured to be arranged closer to the center of the mesa structure than the outer edges of the doping layers 102a and 102b.

The oscillator 200 according to this embodiment will be described with reference to FIG. 7.

The oscillator 200 is an oscillator for oscillating an oscillation frequency fosc = 0.95 THz. In this embodiment, a resonance tunnel diode (RTD) is used as the negative resistance element 201. The RTD used in this example constitutes, for example, a multiple quantum well structure made of InGaAs / InAlAs and InGaAs / AlAs as the active layer 203 on the InP substrate 230. Further, the active layer 203 is configured as a negative resistance element 101 with an electrical contact layer made of n-InGaAs as the doping layers 202a and 202b. As the multiple quantum well structure, for example, a triple barrier structure is used. More specifically, it is composed of a semiconductor multilayer film structure of AlAs (1.3 nm) / InGaAs (7.6 nm) / InAlAs (2.6 nm) / InGaAs (5.6 nm) / AlAs (1.3 nm). Of these, InGaAs is a well layer, and lattice-matched InAlAs and unmatched AlAs are barrier layers. These layers are intentionally left undoped without carrier doping. The active layer 203 having such a multiple quantum well structure is sandwiched between the electrical contact layers of the n-InGaAs doping layers 202a and 202b having an electron concentration of 2 × 10 ¹⁸ cm ^-3 , and is a negative resistance element. It constitutes 101.

In the current-voltage I (V) characteristics of the structure between the electrical contacts, the peak current density is 280 kA / cm ² , and the negative resistance region is from about 0.7 V to about 0.9 V. When the RTD is a mesa structure of about 2 μmΦ, a peak current of 10 mA and a negative resistance of -20 Ω can be obtained.

The patch antenna 204 is a square patch with one side of the conductor 208 having a side of 60 μm, and a BCB (benzocyclobutene, manufactured by Dow Chemical Co., Ltd., ε _r = 2) having a thickness of 3 μm as a dielectric 207 between the conductor 208 and the conductor 209 on the ground side. .4) was placed. An RTD 201 having a diameter of 2 μm is connected between the conductors 208 and 209, and the RTD 201 is arranged at a position shifted by 24 μm in the resonance direction from the center of the conductor 208. The single resonance frequency of the patch antenna 204 is about 1.5 THz, but considering the reactance of the RTD which is the negative resistance element 201, the oscillation frequency fosc of the oscillator 200 is about 0.95 THz.

The conductor 208 is connected to the MIM capacity 2032 via two strip conductors 2031 having a width of 5 μm and a length of 15 μm. The size of the MIM capacity 2032 was set to 10 pF in this embodiment. Wiring 206 including wire bonding is connected to the MIM capacity 2032, and the bias voltage of the negative resistance element 201 is adjusted by the power supply 205.

The negative resistance element 201 has the doping layers 202a and 202b and the active layer 203, and the amorphous layers 210a and 210b are formed on the side walls of the doping layers 202a and 202b, and the amorphous layer 210c is formed on the side walls of the active layer 203. There is.

The active layer 203 (RTD) in this example was formed with a mesa structure of 2 μmφ, and amorphous layers 210a, 210b, and 210c were all formed around the active layer 203 (RTD) with a thickness of 140 nm. Further, the height of the mesa of the negative resistance element 201 was set to 200 nm.

The resistivity of the doping layers 202a and 202b is about 3 × 10 ^-7 Ωm, while the resistivity of 210a and 210b of the amorphous layer is reduced due to the deterioration of crystallinity and the decrease of dopant, and the resistivity is 2. It is estimated to be an increase of up to 3 digits.

As described above, the combined resistance of the doping layers 202a and 202b and the amorphous layers 210a and 210b increases with increasing frequency due to the skin effect. With such a configuration, the oscillation frequency fosc = 0.95 THz is 0.1 Ω or less, whereas the second harmonic (1.9 THz) is 2 to 4 Ω, and the third harmonic (2.85 THz) is 30 to 50 Ω. Is estimated.

Further, in the negative resistance element 201, the dielectric loss increases due to the amorphous layer 210c on the side wall of the active layer 203. The dielectric loss tangent (tan δ) when the amorphous layer 210 is provided on the side wall is estimated to be about 5 times that when the amorphous layer 210c does not exist from the measurement results in the millimeter wave band using an impedance analyzer. With this configuration, the loss of the second harmonic (1.9 THz) and the third harmonic is large with respect to the oscillation frequency fosc = 0.95 THz, and unnecessary oscillation of the second and third harmonics is suppressed. Can be done.

The oscillator 200 according to this embodiment is manufactured as follows.

First, the next layer is epitaxially grown on the InP substrate 230 by a molecular beam epitaxy (MBE) method, an organometallic vapor phase epitaxy (MOVPE) method, or the like. That is, in order, a resonance tunnel diode (RTD) made of n-InP / n-InGaAs, InGaAs / InAlAs, and n-InP / n-InGaAs is epitaxially grown. When an n-type conductive substrate is selected as the InP substrate 230, it may be epitaxially grown from n-InGaAs.

Next, the resonance tunnel diode 201 is etched into an arc-shaped mesa having a diameter of 2.28 μm. EB (electron beam) lithography and dry etching by ICP (inductively coupled plasma) are used for etching. Photolithography may be used.

Subsequently, the side wall of the mesa-shaped resonance tunnel diode 201 is amorphized. The amorphization treatment was carried out using O ₂ plasma to amorphize the resonance tunnel diode 201 from the surface side of the side wall to a thickness of 140 nm so that the diameter of the resonance tunnel diode 201 was 2 μm.

In addition to O ₂ plasma, a dry process such as N ₂ plasma or Ar plasma can also be used.

Further, it is also possible to form a film to be a resistance film instead of amorphizing. In that case, the resonance tunnel diode 201 has a diameter of 2 μm and is formed in a mesa shape, and then a resistance film is formed.

As a component of the resistance film and a film forming method, a CVD method or a CVD method is used in which Si, O, N, F, P, S, etc. are reacted with a low resistance material such as Al, Ti, C so as to obtain a desired resistivity. It can be formed by using a sputter method, a vapor deposition method, or the like.

Subsequently, the ground conductor 209 is formed on the etched surface by the lift-off method.

It is also possible to form the ground conductor 209 before the above-mentioned amorphization treatment. Further, a passivation for protecting the resonance tunnel diode and the amorphous layer on the side wall may be formed. Further, embedding with BCB which is a dielectric 207 is performed by a spin coating method and dry etching, and an upper electrode of a Ti / Pd / Au conductor 208, a strip conductor 2031 and a MIM capacity 2032 is formed by a lift-off method.

Finally, the Bi pattern is formed in the portion to be the resistor 213 by the lift-off method, and the ground conductor and the upper electrode of the MIM are connected to complete the oscillator 200 of this embodiment.

The power to be supplied to the oscillator 200 may be appropriately supplied from the power supply 205 and the wiring 206 via the strip conductor 2031 arranged in the center of the oscillator 200, and usually a bias voltage in the negative resistance region is applied to the bias current. When supplied, it operates as an oscillator.

The oscillator 300 according to this embodiment will be described with reference to FIG.

8 (a) is a perspective view of the oscillator 300, and FIG. 8 (b) is a sectional view taken along the line AA'of the oscillator 300.

In the second embodiment, an example in which the resistance element 314 is arranged in the patch antenna 304 is shown. The resistance element 314 is arranged as a parallel resistance arranged in parallel with the negative resistance element 301 at the position of the node of the high frequency electric field standing in the patch antenna 304.

The resistance element 314 is connected between the conductor 308 and the ground conductor 309 so as to be 20Ω. As disclosed in Patent Document 1, the resistance element 314 can suppress parasitic oscillation caused by the wiring structure.

Further, Example 2 is also an example in which the oscillation frequency fosc is changed with respect to Example 1. The oscillator 300 is an oscillator for oscillating an oscillation frequency fosc = 0.60 THz, and uses the same resonance tunnel diode (RTD) as in Example 1 as a negative resistance element 301 and a patch antenna 304 as a resonator.

The diameter of the resonance tunnel diode (RTD) was formed to be 3 μmφ, and the amorphous layer 310 was formed around the resonance tunnel diode (RTD) to a thickness of 200 nm. Further, the height of the mesa of the negative resistance element 301 was set to 200 nm. In addition, the description of the same structure as that of the first embodiment will be omitted.

The patch antenna 304 is a square patch having a side (L) of 150 μm, and the conductor 308 is connected to a power source (not shown) via one microstrip line 3031 having a width of 5 μm and a length of 38 μm. A 3 μm-thick BCB was placed between the conductor 308 and the conductor 309 as a dielectric 307. An RTD, which is a negative resistance element 301 having a diameter of 3 μm, is connected between the conductors 308 and 309, and the negative resistance element 301 is arranged at a position shifted by 60 μm in the resonance direction from the center of the conductor 308. The single resonance frequency of the patch antenna 304 is about 0.8 THz, but considering the reactance of the negative resistance element 301, the oscillation frequency fosc is about 0.60 THz.

With such a configuration, the combined resistance of the doping layers 302a and 302b and the amorphous layers 310a and 310b is 0.1Ω or less at the oscillation frequency fosc = 0.6THz, whereas it is 3 to 3 at the second harmonic (1.2THz). It is estimated to be 6Ω. Further, in the third harmonic (1.8 THz), it is estimated to be 20 to 30 Ω.

Further, it is estimated that the dielectric loss tangent (tan δ) when the amorphous layer 310c is provided on the side wall is about five times as large as that when the amorphous layer 310c does not exist, as in the first embodiment. With this configuration, the loss of the second and third harmonics becomes large with respect to the oscillation frequency fosc = 0.6 THz, and unnecessary oscillation can be suppressed.

With the configuration of Example 2, both the second and third harmonics and the low frequency parasitic oscillation caused by the wiring structure can be suppressed with respect to the oscillation frequency fosc = 0.6 THz, and the oscillation frequency is stable at a desired oscillation frequency. Oscillation is possible.

(Modification example)
In the above, the triple barrier resonance tunnel diode made of InGaAs / InAlAs and InGaAs / AlAs grown on the InP substrate has been described as the RTD. However, the semiconductor device of the present invention can be provided not only by these structures and material systems but also by a combination of other structures and materials.

For example, a resonance tunnel diode having a double barrier quantum well structure or a resonance tunnel diode having four or more multiple barrier quantum wells may be used. The material system may be GaAs / AlGaAs /, GaAs / AlAs, or InGaAs / GaAs / AlAs formed on the GaAs substrate. Further, it may be InGaAs / AlGaAsSb formed on the InP substrate. Further, it may be InAs / AlAsSb or InAs / AlSb formed on the InAs substrate. Further, it may be SiGe / SiGe formed on the Si substrate.

Further, it may be a combination of at least a part thereof. These structures and materials may be appropriately selected according to a desired frequency and the like.

Although the description is made on the assumption that the carrier is an electron in the present specification, the description is not limited to this, and holes may be used.

The material of the substrate and dielectric may be selected according to the application, and semiconductors such as silicon, gallium arsenide, indium arsenic, and gallium arsenide, and resins such as glass, ceramics, Teflon (registered trademark), and polyethylene terephthalate may be used. You may use it.

Further, the embodiments and examples described above can be replaced or combined with each other.

100 Oscillator 101 Negative resistance element 102 Low resistance semiconductor layer 103 Active layer 104 Patch antenna 105 Power supply 106 Wiring 107 Dielectric 108, 109 Conductor 110 Amorphous layer 112 Resistance 1031 Strip conductor 1032 Capacity

Claims (14)

An oscillator that oscillates terahertz waves
A negative resistance element having a first semiconductor layer, a second semiconductor layer, and an active layer provided between the first semiconductor layer and the second semiconductor layer.
It has a resonator having a first conductor, a second conductor, and a dielectric provided between the first conductor and the second conductor.
The negative resistance element is provided between the first conductor and the second conductor.
A layer is provided between the negative resistance element and the dielectric.
The oscillator is characterized in that the layer is a layer having a resistivity higher than the resistivity of the first semiconductor layer or the second semiconductor layer. The oscillator according to claim 1, wherein the layer is an amorphous layer. An oscillator that oscillates terahertz waves
A negative resistance element having a first semiconductor layer, a second semiconductor layer, and an active layer provided between the first semiconductor layer and the second semiconductor layer.
It has a resonator having a first conductor, a second conductor, and a dielectric provided between the first conductor and the second conductor.
The negative resistance element is provided between the first conductor and the second conductor.
A layer is provided between the negative resistance element and the dielectric.
The layer is an oscillator characterized by being an amorphous layer. The oscillator according to claim 3, wherein the layer has a resistivity higher than the resistivity of the first semiconductor layer or the second semiconductor layer. The oscillator according to any one of claims 1 to 4, wherein the layer has an element constituting the negative resistance element. The oscillator according to any one of claims 1 to 5, wherein the negative resistance element is surrounded by the dielectric. The oscillator according to any one of claims 1 to 6, wherein the negative resistance element has a mesa structure, and the layer is provided on a side wall of the mesa structure. The first semiconductor layer, the active layer, and the second semiconductor layer are laminated in the first direction.
The oscillator according to any one of claims 1 to 7, wherein the layer has a film thickness of 1 nm or more and 500 nm or less in a second direction orthogonal to the first direction. The oscillator according to claim 8, wherein in the second direction, the layer has a film thickness of 5 nm or more and 200 nm or less. The layer includes a first layer provided between the first semiconductor layer and the dielectric, or between the second semiconductor layer and the dielectric, and the active layer and the dielectric. The oscillator according to any one of claims 1 to 9, wherein the oscillator has a second layer provided between the body and the body. The oscillator according to claim 10, wherein the impedance of the first layer is lower than the impedance of the second layer at the oscillation frequency of the oscillator. The oscillator according to any one of claims 1 to 11, wherein the layer has an element constituting the negative resistance element. The oscillator according to claim 12, wherein the element is any one of In, Ga, As, and P. The oscillator according to any one of claims 1 to 13, wherein the resonator is a microstrip resonator having a patch antenna.

Images